The development of internet technologies, especially the development of mobile internet technologies, greatly changes the work and life of people. The creation of social networking applications (such as various instant messaging tools and community networks) makes the interpersonal communication more convenient and diversified. To stabilize and expand a user relationship chain, and to prevent losing social networking application users, various social networking applications usually provide a user recommendation function for recommending to a user other users that the user may be interested in.
A conventional user recommendation method includes a friend recommendation based on a six degree spatial theory and a friend recommendation based on common interest. The friend recommendation based on a six degree spatial theory refers to a theory in that a person can know a stranger via six people at most, and this type of user recommendation method relies on a user relationship chain. The friend recommendation based on common interest refers to a method that recommends users having the same interest, and this type of user recommendation method relies on personal information of the user. However, the conventional friend recommendation method based on a six degree spatial theory requires a complex six-degree relationship calculation process. Further, in the friend recommendation method based on common interest, the establishment of an interest graph model is complex. Therefore, implementation of the conventional user recommendation method is complex, and the recommendation performance and efficiency thereof are not high.